Dear Bernd
I have a question about partitioning of Si during solidification.
Is there any way i can put partitioning coefficient in MICRESS?
I found that in FeCMnAl system, Si is always partioned into liquid phase when i do thermocalc coupling simulation during solidification. I want Si to reside in the growing ferrite.
Thank you
Best regard
Partioning issue
Re: Partioning issue
Hi betleenkim,
The trivial answer to your question is that you can put partitioning coefficients when you use linearised phase diagrams. But perhaps you are thinking about doing so only for the element Si while keeping the thermodynamic description of the other elements - unfortunately this is not possible! If you want to use a mixed linearised-TQ description in MICRESS, this is possible only on a per phase base, i.e. for example "linearised" for ferrite and all its interactions to other phases, and "database" for all the rest.
But there are always ways how things can be achieved somehow:
a) Use a better database (if available)
b) Modify the database. This is easy for free databases (if you know about Calphad, of course!), and even for encrypted databases you can remove elements from sublattices etc.
c) Create a linearised phase diagram from a TQ calculation (linearisation data from .log or .TabLin file). Using the "linTQ" mode in MICRESS, you can directly use this output format. Then, you just modify the c0 data for Si. Problem: you can only chose the linearisation data for one temperature. If it changes too much with temperature, the only chance is to make several "restart" calculations with modified linearisation data for different temperature regions...).
d) You could check out if by chance there is another element in the database which behaves like you want it for Si
Bernd
The trivial answer to your question is that you can put partitioning coefficients when you use linearised phase diagrams. But perhaps you are thinking about doing so only for the element Si while keeping the thermodynamic description of the other elements - unfortunately this is not possible! If you want to use a mixed linearised-TQ description in MICRESS, this is possible only on a per phase base, i.e. for example "linearised" for ferrite and all its interactions to other phases, and "database" for all the rest.
But there are always ways how things can be achieved somehow:
a) Use a better database (if available)
b) Modify the database. This is easy for free databases (if you know about Calphad, of course!), and even for encrypted databases you can remove elements from sublattices etc.
c) Create a linearised phase diagram from a TQ calculation (linearisation data from .log or .TabLin file). Using the "linTQ" mode in MICRESS, you can directly use this output format. Then, you just modify the c0 data for Si. Problem: you can only chose the linearisation data for one temperature. If it changes too much with temperature, the only chance is to make several "restart" calculations with modified linearisation data for different temperature regions...).
d) You could check out if by chance there is another element in the database which behaves like you want it for Si
Bernd